Four companies. One week of earnings. One number that changes everything.
Microsoft, Alphabet, Meta and Amazon are now expected to spend nearly $725 billion combined this year to fuel their AI buildouts. That is not a budget line. That is a civilizational bet. The largest coordinated capital deployment in corporate history, all pointed at the same target.
| Company | 2026 AI Capex |
|---|---|
| Amazon | $200 billion |
| Microsoft | $190 billion |
| Alphabet | $180 to $190 billion |
| Meta | $125 to $145 billion |
| Total | ~$725 billion |
Every dollar in that table needs a physical home. Data centers. Power contracts. Cooling systems. Fiber. The question is where all that infrastructure actually comes from.
The Numbers Behind the Bet
Microsoft reported Q3 FY2026 revenue of $82.9 billion, up 18% year over year, beating analyst expectations. Azure revenue grew 40%, ahead of the 39% analysts had penciled in. Microsoft Cloud revenue hit $54.5 billion, up 29%. AI revenue surpassed a $37 billion annual run rate, up 123% year over year. Capital expenditure for Q3 alone was $31.9 billion. Full year capex is now guided at $190 billion, far above the $154 billion analysts had expected. Despite beating on every metric, Microsoft stock fell more than 3% in after hours trading. The market is not rewarding results anymore. It is demanding proof that $190 billion in spending will generate returns faster than the depreciation clock.
Meta announced $19.84 billion in capital expenditure in Q1 2026 and raised its full year forecast to $125 to $145 billion, up $10 billion at both ends from prior guidance. When Meta's CFO was asked about future buybacks on the earnings call, her answer was direct. The highest priority is positioning the company as a leader in AI. Meta stock fell about 7% in after hours trading after missing on user growth, partly attributed to internet disruptions in Iran.
Amazon expects to invest $200 billion in capital expenditures across its business in 2026. AWS grew 24% in the most recent quarter, its fastest growth in 13 quarters. Andy Jassy called it a seminal opportunity alongside chips, robotics and low earth orbit satellites. Amazon spent $43.2 billion in Q1 alone as AWS accelerated to 29% year over year growth.
Alphabet doubled its AI spending compared to last year, now guiding between $180 and $190 billion. Google Cloud grew 63% in Q1, one of the fastest growth rates in the company's history. Alphabet stock jumped 7% after its report, the clear winner of earnings week among the four hyperscalers.
These are not projections. These are commitments already in motion. Data centers are being built. Power contracts are being signed. Cooling systems are being installed. The physical infrastructure of the AI economy is going into the ground right now.
The Infrastructure Hiding in Plain Sight
Every one of those data centers needs three things. Power, cooling and connectivity.
Power is the bottleneck. The US grid was not built for this. Utility companies are warning that AI data centers are pushing regional grids to their limits. New nuclear plants take 10 years to permit. New gas plants take 3 to 5 years. Solar and wind cannot deliver the consistent baseload these facilities need.
Bitcoin miners figured this out a decade ago.
The infrastructure that Bitcoin miners spent years building, the substations, the high voltage lines, the cooling arrays, the relationships with power companies, was never just for mining. It was the only private sector buildout of serious energy infrastructure that happened outside of traditional utility planning. Those miners went to places nobody else wanted, negotiated power deals nobody else bothered with, and built physical infrastructure that now happens to be exactly what AI needs.
Gensyn listed this week on Binance, Coinbase and Kraken, built on exactly this thesis. It combines global computing power into a single open network for machine learning. The protocol connects idle compute the same way Bitcoin connected idle hashing power. The architecture is familiar because the problem is the same.
What This Means for Bitcoin
The $725 billion AI buildout does something concrete for Bitcoin that most analysts have not priced in yet. It validates the energy infrastructure thesis entirely. Every gigawatt of power that goes into an AI data center is a gigawatt that had to be sourced, stabilized and delivered. Bitcoin miners have been solving exactly that problem in exactly the same locations for years.
The miners that survive the next 18 months will not just be mining Bitcoin. They will be leasing compute, providing grid stabilization services and running inference workloads during off peak hours. The economics of mining are converging with the economics of AI infrastructure faster than the market has priced in.
Bitcoin sits at the intersection of power, cooling and compute density. That is exactly where $725 billion is flowing.
The Free Cash Flow Warning
Not everything in these earnings is straightforward. Amazon is projected to turn negative free cash flow this year. Meta's free cash flow is expected to drop almost 90% according to Barclays analysts. Microsoft free cash flow came in at $15.8 billion for the quarter, down significantly as capex consumed the difference. Gross margin compressed to 67.6%, the narrowest since 2022, as data center depreciation costs mounted.
Meta stock fell 7% despite beating on revenue. Microsoft stock fell 3% despite beating on every metric. The market is not reacting to results. It is reacting to the size of the bill and the uncertainty of the return timeline.
These companies are burning cash at a rate that would concern any traditional investor. The only reason markets are not panicking is because the revenue growth is validating the spend in real time. If growth slows, the repricing will be fast and severe.
What to Watch Going Forward
Watch power purchase agreements. When Microsoft, Amazon or Google sign a major energy deal in a region where Bitcoin miners operate, the thesis is playing out in real time.
Watch GPU allocation. Nvidia cannot produce enough chips to meet demand. Any company that controls compute infrastructure before the supply catches up holds a structural advantage. Bitcoin miners with existing power and cooling are first in line.
Watch distributed compute protocols. The AI compute category raised over $221 million across four listings this week alone. The market is pricing in a future where distributed compute is as valuable as centralized cloud.
Watch Bitcoin miner earnings in Q2. The companies that have pivoted toward AI compute hosting will start showing it in their revenue mix. That is the moment the market connects the dots between the Bitcoin infrastructure thesis and the $725 billion AI bet.
Watch Meta specifically. Meta has the most aggressive spending setup and faces the most pressure to show returns. If Meta's AI products start generating measurable revenue in Q2, the entire narrative shifts from "is AI worth it" to "how much more should we spend."
Watch Azure guidance for Q4. Microsoft guided Azure growth of 39% to 40% for the next quarter. If it hits the high end, the $190 billion capex gets justified fast. If it misses, the repricing starts.
The $725 billion is already committed. The infrastructure is already being built. The only question left is who owns the rails it runs on.
BitBrainers. We check the facts so you don't have to.
